In the traditional enterprise data protection scenario, the backup window (e.g., the time to perform a backup) is decided by the backup administrator based on peak usage trends on the production environment. With an enterprise level data as the backup interest, protection of such critical data becomes the priority. Meanwhile, the peak load balance on the production environment is required as well. Consequently, the backup administrator would need to monitor the resource usage continuously and adjust the backup window to mitigate the peak load on the production server. Also, during the backup phase, the cumulative load from both backup application and native production applications consume important system resources causing performance bottleneck.
With the increase in the scale of production, of native applications and in the number of production servers, the system resource monitoring could become significant overhead to the backup/system administrator. Also, such human decisions could be prone to inaccuracies, and may not always lead to the optimal utilization of system resources.